Publicación:

Traductor de video y señas (Travid)

Resumen

Currently, job offers need systems engineers of high quality with the skills necessary for software development, database among other knowledge that incorporate this career. It is observed in the system engineering program of the Universidad Cooperativa de Colombia, how the students perceive day at day to the society immersed on the technology thanks to the professional formation, and the study’s plan which has being acquired during the last five years. This has helped to simplify many homework many times are think as complex. But in the area of inclusions (SDG 10), and especially in the area of people with hearing disabilities, exists important spaces for generate developments that contribute on the inclusion of this population. This problem is presented in a sociopolitical context that currently exists, where people with this disability by lack of communication channels are not taken into account and are exiled from social development where they have to be included. It has been seen that on the actual society the people with hearing disabilities are very little included in social development, for not having means or alternatives to interact with hearing people. For mitigate part of this situation we propose an application, through image recognition will translate sign language to text, based on the image recognition algorithms that use Matlab software. 160 tests were performed for each of the five vowels, these tests consist of twenty different parts, in which each of them varied different aspects such as far, near, with white glove, with the left and right hand, this sampling was reflected in different graphs in order to have percentage results of success. Some important aspects to consider are that the vowel A turned out in a variation different to the other vowels, this variation consists that our applicative could recognize much easy this vowel followed by the vowel E, because both are similar and in other minor cases it recognize the I, O and U, the vowel E identify with a good result showing all the 160 tries of this one, in the vowel I identify a variation between I, E and U, where the major was I, in the vowel O the results were I, E and O where the major one was the vowel O and the vowel U show us the vowels E, I, A and U where the U was the major translation.